libdrm_radeon: Optimize copy of table to cs buffer with specialized call.
Using this call in OUT_BATCH_TABLE reduces radeonEmitState cpu usage from 9% to 5% and emit_vpu goes from 7% to 1.5%. I did use calgrind to profile gears for cpu hotspots with r500 card. Signed-off-by: Pauli Nieminen <suokkos@gmail.com>main
parent
a474fd978c
commit
8c43b79b21
|
@ -201,6 +201,15 @@ static inline void radeon_cs_write_qword(struct radeon_cs *cs, uint64_t qword)
|
|||
}
|
||||
}
|
||||
|
||||
static inline void radeon_cs_write_table(struct radeon_cs *cs, void *data, uint32_t size)
|
||||
{
|
||||
memcpy(cs->packets + cs->cdw, data, size * 4);
|
||||
cs->cdw += size;
|
||||
if (cs->section) {
|
||||
cs->section_cdw += size;
|
||||
}
|
||||
}
|
||||
|
||||
static inline void radeon_cs_space_set_flush(struct radeon_cs *cs, void (*fn)(void *), void *data)
|
||||
{
|
||||
cs->space_flush_fn = fn;
|
||||
|
|
Loading…
Reference in New Issue